Generative adversarial network based on domain adaptation for crack segmentation in shadow environments
Yingchao Zhang, Cheng Liu
Abstract
Precision segmentation of cracks is important in industrial non-destructive testing, but the presence of shadows in the actual environment can interfere with the segmentation results of cracks. To solve this problem, this study proposes a two-stage domain adaptation framework called GAN-DANet for crack segmentation in shadowed environments. In the first stage, CrackGAN uses adversarial learning to merge features from shadow-free and shadowed datasets, creating a new dataset with more domain-invariant features. In the second stage, the CrackSeg network innovatively integrates enhanced Laplacian filtering (ELF) into high-resolution net to enhance crack edges and texture features while filtering out shadow information. In this model, CrackGAN addresses domain shift by generating a new dataset with domain-invariant features, avoiding direct feature alignment between source and target domains. The ELF module in CrackSeg effectively enhances crack features and suppresses shadow interference, improving the segmentation model's robustness in shadowed environments. Experiments show that GAN-DANet improves the crack segmentation accuracy, with the mean intersection over union value increasing from 57.87 to 75.03, which surpasses the performance of existing state-of-the-art domain adaptation algorithms.